Mickmumpitz
patreon
Shoot ENTIRE MOVIES with this FREE AI WORKFLOW! [NEW VIDEO + FREE WORKFLOWS]
🕑 Added 2025-05-01 11:41:43 +0000 UTC👇 Download the FREE workflows below.
👉 Download the ADVANCED workflows & example files here: [COMING VERY SOON]
📝A groundbreaking free & local AI video model is changing filmmaking, giving creators access to techniques once reserved for major studios. Using the open-source Wan2.1 video model, you can now replace actors, transform visual styles, generate new shots from text, and even change camera movements in existing footage—all without expensive equipment or software subscriptions. To demonstrate these capabilities, we filmed an entire Kafka-inspired short film on a smartphone, transforming myself into a giant beetle who wakes up one morning to find his life completely changed.
Our approach combines several AI techniques, including strategic use of ControlNets that guide the video generation process. By extracting pose data from original footage while preserving background elements, we achieved seamless character transformation while maintaining consistent camera movement. To ensure our beetle character looked the same across all shots, we trained a custom LoRA model on carefully curated reference images. The results were stunning—professional-quality visual effects generated entirely through free, accessible tools.
We've packaged everything into easy-to-use ComfyUI workflows (tested on an RTX 2070 8GB VRAM & RTX4090 24GB VRAM) that anyone can download and implement. Whether you want to create surreal transformations, replace actors in existing footage, or even automatically remove objects from videos (outperforming Adobe's expensive Content-Aware Fill), our free toolkit puts Hollywood-caliber visual effects at your fingertips. So whether you're looking to level up your AI filmmaking skills or just want to turn yourself into a cute anime girl, we've got you covered—and make sure to watch till the end for the full Kafka-inspired short film!
1. Switch out the character with: 250501_MICKMUMPITZ_FLUX_INPAINTING_V01
💡Guide: https://mickmumpitz.ai/guides/basic-flux-inpainting-workflow-easily-replace-characters
2. Transform the video with:
250501_MICKMUMPITZ_WAN2_1_StartFrame+CN_SMPL_V01
💡Guide: https://mickmumpitz.ai/guides/replacing-actors-in-videos-with-comfyui-and-wanvideo
3. Remove objects and subjects: 250501_MICKMUMPITZ_WAN2_1_CLEAN_PLATE_CREATOR_v01
💡Guide: https://mickmumpitz.ai/guides/remove-objects-from-videos-easily
4. Start Frame + End Frame workflow: 250501_MICKMUMPITZ_WAN2_1_StartFrame+EndFrame_v01
💡Guide: https://mickmumpitz.ai/guides/wan-startendframe-guide
🆕 CHANGE LOG
05/15/2025. Just updated the workflow to work with the new ComfyUI frontend. Use v02!
ComfyUI version: 0.3.34
ComfyUI frontend version: 1.18.10
![Shoot ENTIRE MOVIES with this FREE AI WORKFLOW! [NEW VIDEO + FREE WORKFLOWS]](https://img5.patrefans.com/storage/4/yu/zy/563145-019e8569-da56-771d-a3ee-d2b4f54f2fc9.jpg)
Comments
Caro
Hey, to get a reply to your question, please join the Discord server and post it in the "need-help" channel. https://support.patreon.com/hc/en-us/articles/212052266-Getting-Discord-access
Caro
Hey, to get a reply to your question, please join the Discord server and post it in the "need-help" channel. https://support.patreon.com/hc/en-us/articles/212052266-Getting-Discord-access
Drawgate, Inc
Hi, I've downloaded the 250501_MICKMUMPITZ_WAN2_1_StartFrame+CN_SMPL_v02, and it is not giving any errors; however, the output of the video is only for the skeleton of the controlnet model not a full video. I followed the inpainting Json and it works. What went wrong and how can I get the video output correctly rendered instead of getting the skeletons. Please help. Thank you so much in advance.
anirban borthakur
Successfully uploaded Image and connected the image to preview bridge. Used maked editor and ran it. And nothing happens at the Sampler, it remains blank. Why?
HarryBoBarry
"Download the ADVANCED workflows & example files here: [COMING VERY SOON]" i signed up just for this
Doğukan AYAR
CLIPTextEncode ERROR: clip input is invalid: None If the clip is from a checkpoint loader node your checkpoint does not contain a valid clip or text encoder model. has anyone solved this problem?
